Slashdot Mirror

Take a look at the kernel apis documented here: http://msdn.microsoft.com/library/default.asp?url= /library/en-us/memory/base/memory_management_funct ions.asp

This ought to give you a better idea of the kind of calls they're probably making improvements to.

Maybe the new thing for Vista is just that the low-fragmentation heap is the default mode.

From their faq on the subject:

Microsoft will offer a minimum of 10 years of support for Business and Developer products. Mainstream support for Business and Developer products will be provided for 5 years or for 2 years after the successor product (N+1) is released, whichever is longer. Microsoft will also provide Extended support for the 5 years following Mainstream support or for 2 years after the second successor product (N+2) is released, whichever is longer. Finally, most Business and Developer products will receive at least 10 years of online self-help support.

The statements about 2k and xp are entirely true. ... [Microsoft] lived up to their goals for xp and 2k.

Excuse me!? :

While it may be true that their goals for Windows 2000 did include "to improve security and safety for all our customers", it was definately not their primary goal, and they failed to deliver real improvements in this area. Anyone with real security credentials would have told them ActiveX needed to go, but that meant they would loose their stranglehold on the Intranets. Internet Explorer exploits were still rampant until XP SP2, and ActiveX still has not been removed from the product, nor does it look like IE 7 will finally kill it.

"Windows XP is the most secure and dependable operating system we have ever produced." (source)

A lot of folk consider Windows 2000 to be more dependable than Windows XP, and DOS was more secure; back when Novell built the TCP/IP stack for their operating system.

What a pity they neglected to compare XP with operating systems produced outside of Redmond -- maybe with the bar set a bit higher in the first place, they may have been able to avoid taking a year out to produce a secure version of XP: SP2.

Maybe with Longhorn, Microsoft will finally achieve the goals they set for themselves in the late 90's.

Process memory allocation != heap allocation. The smallest amount of memory the kernel can allocate to a process is one page, which is 4096 bytes on the i386 architecture. Linux operates exactly the same way. To request more memory be committed and mapped in a process, the NtAllocateVirtualMemory syscall is used. 4096 bytes is much too large for most general purpose allocations, so a heap structure is used to further divide up the memory size. The standard heap code (used by the Win32 heap functions and Microsoft's C runtime) resides in ntdll.dll, and are executed entirely in user mode. When the heap code needs more pages to expand the heap, it surely does call NtAllocateVirtualMemory, but most allocations are done on a much smaller scale.

XP introduced (and had backported to 2000) the Low Fragmentation Heap option that uses presized buckets to reduce long term fragmentation. That's the only thing I can think of that they could be talking about. It's part of the standard heap code in ntdll running in user mode.

Microsoft certainly didn't rewrite "large parts of the kernel." If they were going to do something on that scale, you'd think they would ditch (or at least marginalize) the awful Win32 subsystem, but they aren't. This is one of the worst attempted technical articles I've seen in a while. Besides, the kernel never was the weak point in Windows NT's (yes, Vista is still NT) security model, or as an OS in general.

Process memory allocation != heap allocation. The smallest amount of memory the kernel can allocate to a process is one page, which is 4096 bytes on the i386 architecture. Linux operates exactly the same way. To request more memory be committed and mapped in a process, the NtAllocateVirtualMemory syscall is used. 4096 bytes is much too large for most general purpose allocations, so a heap structure is used to further divide up the memory size. The standard heap code (used by the Win32 heap functions and Microsoft's C runtime) resides in ntdll.dll, and are executed entirely in user mode. When the heap code needs more pages to expand the heap, it surely does call NtAllocateVirtualMemory, but most allocations are done on a much smaller scale.

XP introduced (and had backported to 2000) the Low Fragmentation Heap option that uses presized buckets to reduce long term fragmentation. That's the only thing I can think of that they could be talking about. It's part of the standard heap code in ntdll running in user mode.

Microsoft certainly didn't rewrite "large parts of the kernel." If they were going to do something on that scale, you'd think they would ditch (or at least marginalize) the awful Win32 subsystem, but they aren't. This is one of the worst attempted technical articles I've seen in a while. Besides, the kernel never was the weak point in Windows NT's (yes, Vista is still NT) security model, or as an OS in general.

Hum, finally a good comment on this. You know, you have quite a few alternatives in x86/amd64 operating systems, even from the Great Satan itself!

This kind of attitude has caused the largest problem with the public's perception of software development.

"I got it to work, so I must be a programmer. See it isn't that hard."

I have worked at a company that used one of those 4GLs and actually had subject matter experts writing code rather than driving requirements. Some of the constructions that got put into the core product were very hard to maintain and caused certain customer driven customizations and enhancements to be either impossible or extremely difficult (one of the selling points was that you could customize the software somewhat to fit your business model).

This attitude was also a big cause of the dot bomb. If you could spell computer you could get a "software engineering" job.

These people don't know the history of computing and so end up re-writing things that have already been produced and are well known.

Of course I've also seen this kind of thing work well for Microsoft as they actually brag about SMP (Symmetric Multi-Processing http://en.wikipedia.org/wiki/Symmetric_multiproces sing) and CLR (Common Language Runtime http://msdn.microsoft.com/netframework/programming /clr/default.aspx).

DEC had asymetric processing with their VAX 782 processor in the early 80's. If you insist on Intel hardware, we had multiple 80186 and 80286 processors sharing memory in one enclosure running our business in 1985 (80186 and 80286 processors are 8086 processors with extended instruction sets) running a multi-processor predecessor to MS-DOS called MP/M (CP/M, the single processor version is what MS-DOS was based on).

Common Runtime Libraries (RTL) were the way everyone else has always done it as far as I can tell. You called the run-time library or operating system the same way from each language (the syntax was different, but the resulting signature was the same). The compiler took care of making the linkages correct. Microsoft finally normalizes their languages to use a common RTL and pretends it's a totally new concept.

Reason #1 doesn't apply, it costs 250$ to submit a driver for signing : https://winqual.microsoft.com/download/WHQLPOLICY. doc

Reason #2 doesn't apply for the same reason

Reason #3 Eh, simply play non-DRMed content then, playing it at low-quality is better than not being able to play it on your Linux box or other

Reason #4 I call that jaleousy

Why limit yourself to Windows Vista!?

Windows Me

"Windows Me: PC Health Features Keep PCs Stable, Secure and Reliable -- and Take the Frustration Out of Computing for Home Users" (source)

Windows 2000

"Our primary goal is to improve security and safety for all our customers -- consumers and businesses, regardless of size -- through a balance of technology innovation, guidance and industry leadership," Gates said. "We're committed to continued innovation that addresses the threats of today and anticipates those that will undoubtedly emerge in the future." (source)

Windows XP

"Windows XP is the most secure and dependable operating system we have ever produced." (source)

Windows Vista

"In Vista, it should be much more difficult for unauthorized programs (like Viruses and Trojans) to affect the core of the OS and secretly harm your system." (source)

Why limit yourself to Windows Vista!?

Windows Me

"Windows Me: PC Health Features Keep PCs Stable, Secure and Reliable -- and Take the Frustration Out of Computing for Home Users" (source)

Windows 2000

"Our primary goal is to improve security and safety for all our customers -- consumers and businesses, regardless of size -- through a balance of technology innovation, guidance and industry leadership," Gates said. "We're committed to continued innovation that addresses the threats of today and anticipates those that will undoubtedly emerge in the future." (source)

Windows XP

"Windows XP is the most secure and dependable operating system we have ever produced." (source)

Windows Vista

"In Vista, it should be much more difficult for unauthorized programs (like Viruses and Trojans) to affect the core of the OS and secretly harm your system." (source)

Why limit yourself to Windows Vista!?

Windows Me

"Windows Me: PC Health Features Keep PCs Stable, Secure and Reliable -- and Take the Frustration Out of Computing for Home Users" (source)

Windows 2000

"Our primary goal is to improve security and safety for all our customers -- consumers and businesses, regardless of size -- through a balance of technology innovation, guidance and industry leadership," Gates said. "We're committed to continued innovation that addresses the threats of today and anticipates those that will undoubtedly emerge in the future." (source)

Windows XP

"Windows XP is the most secure and dependable operating system we have ever produced." (source)

Windows Vista

"In Vista, it should be much more difficult for unauthorized programs (like Viruses and Trojans) to affect the core of the OS and secretly harm your system." (source)

You didn't look carefully. wind 32 classes. I think you are confusing two things:

1) Whether the functions themselves are organized in an OO way
2) Whether the languages required you to be OO in your thinking

Win16 was OO it was written in organized C++. OTOH VB 3 or whatever was not OO by any stretch. Do you see the difference? This actually pretty SOP for Microsoft they tend not to push people towards better ways of working but rather build products which fit people's percieved desires.

yep

It's not that easy, there is a download required for Windows 2000 and pre-SP2 XP machines. See the MS help article on it.

This post is acronym laden, so I've spelled out the meanings in most cases.

First off, there is an alternative to IPSec that also ships with Windows - PPP/PPTP (Point-to-Point [Tunneling] Protocol), but since I'm nearly unfamiliar with it, I'll let others discuss it, if necessary.

    Firewall ports need to be opened for UDP 50 (ESP - Encapsulating Security Payload) and 500 (ISAKMP - ISA's Key Management Protocol, sometimes also called IKE - Internet Key Exchange) and if there's NAT (Network Address Translation) transversal anywhere in the system (i.e. any routers that redirect traffic to the 192.168.x.x or 10.x.x.x "internal use" domains such as those used by DHCP clients [including wireless]), you will also need UDP port 4500 (ipsec NAT-T).

    I'm not sure if MS uses AH (Authentication Header) or just re-encapsulates at the firewall, but judging by their web page, I doubt it's a problem. In a nutshell, AH was designed by IPv6 people that think NAT is an unnecessary evil and that all machines need to be uniquely identified. In some ways their ideals are good (security guarantees the originating host), but I think the world is getting too paranoid about "Big Brother" for this to be realistic (all items downloaded can be uniquely identified to a machine - nice for the feds, bad for privacy). Basically, AH breaks at the router in NAT. The router repackages the packet to the real host outside the firewall and AH identifies the packet as tampered with. There are workarounds to this, but I don't believe the discussion is necessary.

    You're also making the big assumption that these people are computer literate, as well, and judging by the post and their choice of hardware, I'm assuming not (Pelco is supposed to be easy to use - heck, my ex-boss's computer illiterate husband used one for security at her home, though she was somewhat computer literate [if you count Excel/Powerpoint goddess as computer literate]). Detailed instructions on the step-by-step requirements or more likely, a setup program to install and configure this will be required.

It doesn't matter which version of Windows you run (okay, not necessarily with '95, '98 or something even more ancient) you can install the same .exe file and run it.

Through Microsoft's Fresh Start program for schools, all donated PCs can get a NEW copy of Windows. For FREE (as in beer). All you have to do is prove they were donations (which I expect they were and you can) and you can get a new copy of the OS for each machine.

http://www.microsoft.com/education/FreshStart/Fres hStart.asp

So why did Dell refuse -- no matter how we tried to word the question -- to admit that they really had moved a bit further toward offering Linux on the desktop?

Or, better still, why doesn't Dell just start offering one Linux distribution as an option on their complete desktop line?

The reason is that, when push comes to shove, "Dell recommends the use of Windows XP Professional" on its desktops.

So I, for one, think that the real reason Dell keeps the Linux desktop at arm's reach is that it doesn't want to tick off Microsoft.

The reason is that Dell and others won't receive the kickbacks, oops, marketing funds provided by Microsoft if they advertise Linux or provide other non-Microsoft software on a Windows desktop (Firefox, openoffice, winamp etc.).

From the article here: ahref=http://www.usdoj.gov/atr/cases/f205700/20575 1.htmrel=url2html-6618http://www.usdoj.gov/atr/cas es/f205700/205751.htm>
Pursuant to the MDA, Microsoft provides marketing funds -- in the form of discounts on the price the OEM pays for each copy of Windows -- to OEMs whose print advertisements and websites promote Microsoft's operating systems in a manner specified by Microsoft.

From Microsoft, http://www.microsoft.com/uk/partner/sales_and_mark eting/partner_initiatives/marketing-fund/
From the 1 February 2006 you may qualify for an important new benefit from Microsoft: the System Builder Cooperative Marketing Fund. If you qualify you can accrue funds that can be used to reimburse eligible marketing activities. This flexible funding is offered to Microsoft Partners for every Windows® XP System Builder licence you buy for: Windows XP Professional, Windows XP Media Center Edition, Windows XP Professional 64-bit, and Windows XP Home Edition.

Enjoy,

Why not upgrade to Windows 2000??

http://www.microsoft.com/education/freshstart/fres hstart.asp

Teaching people the concepts of where to find the usual supects in an application: file,edit,view, configuration and how to save as, cut, copy and paste...etc, is far more useful in real life, because they may be using a real world application they've never seen before.

I'm a K-12 network administrator, so I know where you are coming from in terms of budgetary concerns. However, I encourage you to furthur investigate licensing options before you consider a switch of operating systems. For instance, you can buy a retail box acadmic version of XP Pro with SP2 for less than $70 (Microsoft SKU E85-02670). Other products are also available in academic pricing, like Office 2003.

If you are more forward thinking, you could enter into a School Agreement with Microsoft in which you could acquire volume licensing (greater than 300 units) at an even bigger discount (possibly around $30 per workstation). You could calculate a cost here.

One other thing to note, is that if you can solicit computer donations to your school system, either from parents or community members, you are eligible to free (as in beer) Windows 2000 licensing for such computers. This is called the Microsoft Fresh Start program for K-12.

Also check out K12Computers.com. You can occasionally find some really cheap surplus Dells sold only to K-12.

I'm a K-12 network administrator, so I know where you are coming from in terms of budgetary concerns. However, I encourage you to furthur investigate licensing options before you consider a switch of operating systems. For instance, you can buy a retail box acadmic version of XP Pro with SP2 for less than $70 (Microsoft SKU E85-02670). Other products are also available in academic pricing, like Office 2003.

If you are more forward thinking, you could enter into a School Agreement with Microsoft in which you could acquire volume licensing (greater than 300 units) at an even bigger discount (possibly around $30 per workstation). You could calculate a cost here.

One other thing to note, is that if you can solicit computer donations to your school system, either from parents or community members, you are eligible to free (as in beer) Windows 2000 licensing for such computers. This is called the Microsoft Fresh Start program for K-12.

Also check out K12Computers.com. You can occasionally find some really cheap surplus Dells sold only to K-12.

I'm a K-12 network administrator, so I know where you are coming from in terms of budgetary concerns. However, I encourage you to furthur investigate licensing options before you consider a switch of operating systems. For instance, you can buy a retail box acadmic version of XP Pro with SP2 for less than $70 (Microsoft SKU E85-02670). Other products are also available in academic pricing, like Office 2003.

If you are more forward thinking, you could enter into a School Agreement with Microsoft in which you could acquire volume licensing (greater than 300 units) at an even bigger discount (possibly around $30 per workstation). You could calculate a cost here.

One other thing to note, is that if you can solicit computer donations to your school system, either from parents or community members, you are eligible to free (as in beer) Windows 2000 licensing for such computers. This is called the Microsoft Fresh Start program for K-12.

Also check out K12Computers.com. You can occasionally find some really cheap surplus Dells sold only to K-12.

Microsoft agrees that .NET is a framework.

I agree with GP, I do not know why people continue to *believe* that typing cryptic commands to the computer is easy, again it may be straight forward when YOU KNOW HOW but it is not easy to INFERE when you do not know the commands.

for one, what THE FUCK does eselect opengl set nvidia means? that does not makes any sense, again what does emerge nvidia-kernel means??

Give a doctor those instructions and they will just unplug the cord and go to read a book. You have to have a very narrow perspecitve to think like that. One of the main goals of software usuability is to allow people to INFERE by themselves how to use something with their prior knowledge.

People know how to search in google. And if they need to do something they will search in google.

I will take the time to give you a step by step example.
Lets imagine our hypothetical Jane average user who has returned from London and has her digital camera full with pictures, now she wants to make an album on her computer. She has two different choices:

google "photo album windows XP"
or
google "photo album linux".

In the first case, if she has Windows, the first 3 google search results are:
1. Microsoft Windows XP - Make a photo album on your computer
2. Picasa
3. Digital Photography - Reviews and free downloads at Download.com

The frist option is a page containing detailed step by step instructions on how to make a Windows Folder and use it as a photo album.
The second option is Picasa which will direct you to download and install the program by clicking "open" (after that, you know it is Next/Next/Next).
The third option will direct you to download.com, which provides a list of photo album applications like FlipAlbum (ha, my father bought that IIRC).

Now, with our Linux Jane, what do we have as the 3 first results:
JAlbum - free web photo album software and photo gallery software
Corel Corporation - Home of CorelDRAW, WordPerfect, Paint Shop Pro ...
Linux Online - Application: Web Photo Album

The first one is the closer to an answer, it takes you to Jalbum, unfortuantely you must register with some personal data (so much for the Open Source karma uh?) and you may need to download and install Java runtime.

If our Jane have not run yet, she will download the .bin, and then she may try to open it (supposing that she is using fedora core 4 with Gnome she may double click or right click and open, or just click open on the firefox download window) and then a big text full big X error window will appear saying: "Cannot open JAlbuminstall.bin".

Of course, she should know better as to change the rights to executable in the properties windows and then after that, open a terminal and run it... but that was just too much, so she went to the second option... ... Which takes her to Corel corporation, and after wandering around she finds Corel Photo Album, and after clicking on the "Try It" link she may create an account and download the program just to see that it is an EXE, whoops, not for Linux (she might even read the System Requirments before doing all that...).

So, she goes to the third option, it seems good:

Web Photo Album automatically generates photo albums on the fly from directories containing your favorite photos. Supports captions (including HTML tags), definable page sizes, forward and reverse preview, and index pages. All preview and inde

The official Microsoft Windows Vista web site indicates there are only 5 versions. -lotd

• Windows 2000 Pro (I don't think there was a home)
• Windows 2000 Server,
• Windows 2000 Advanced Server
• Windows 2000 Datacenter Server.

6 versions of Vista is getting a little carried away. But it's just a continuation of the theme they've already established. I expect there will be 3 versions of smelly lobotomised OS's, one version most use, and maybe acouple with features no one wants or cares about

btw I don't know where the BBC sourced their "6" from but MS seems to bthink there are five versions

The five different versions of Windows Vista are designed to fit the way you intend to use your PC:

how does Apple deal with bugs in Mac OS 8 and Mac OS 9?

MacOS 8, I don't know. My guess is that it's not supported anymore, so it doesn't get fixed.

As for Microsoft, they don't fix bugs in NT 4 and Windows 95 anymore. Windows 98 and ME will have their support dropped in five months, even though there's still *millions* of installed systems out there.

In any event, security is not based on the number of worms and/or viruses out there for a specific platform.

Looks like you got it -- Vista prompts you for permission to do administrative stuff, even if you're logged on as administrator, rather than granting it by default. Took them a few years, but they finally got that part right.

As for better security from a MS OS, there's always Singularity, but it's several years away.

http://www.microsoft.com/presspass/press/2006/feb0 6/02-26WinVistaProductsPR.mspx

But, Doctor Evil, that already happened.

There is a thing called a "re-gifter". I think it was coined on Seinfeld, but the concept has been around forever. At least since the invention of the fruitcake.

Why get your news from the BBC when Microsoft released this information yesterday?

Well here are a couple of good technical descriptions here and here. Basically for Multicasting to work all that is needed is for the routers between yourself and the streaming server to support Multicasting. The key question is what percentage of the routers out there support Multicasting ?

The answer is not many. Many of the routers out there do support Multicasting however this feature is disabled ironically to reduce the work load on the router. All multicasting does is shift the work load from the server to the router, which does not scale for the exact same reasons unicast servers don't.

I do think multicasting is probably the way to go, but it's probably quite some time away. When optical fibre networks and optical routers start to pentrate into the home market, then these technologies will become more realistic on the internet, until then.... enjoy your Fox box.

instead of a thumbstick controller on a gamepad, they don't include a thumb-sized trackball

Back to mice as a control device ... I think back to Freelancer. Now, most space combat games have been best played with a joystick (with a hat if the game let you slide to the side like Descent did (and though you were underground, it still played like a space combat game)), but Freelancer was different. It let you control your spaceship with a mouse in a reasonable way in an arcade-like setting, and it was wonderful -- all because they decided that the weapons were on a controllable turret rather than just pointing straight ahead from your ship (and since your ship is massive, it can't be turned as quickly as your mouse moved.) It just worked ...

Seriously, if C bothers you that much, don't use it.
Seriously ? If I could avoid using C, I would. But there is a perverse logic by which every "serious" application or library seems to be doomed to be written in C or, in the best case, C++, regardless of the application. Don't take me wrong. C or C++ is probably the best language for low-level, embedded and/or kernel development. But not for writing, say, Mozilla (according to my measurements nearly 2/3 of the native code of Mozilla is actually spent mapping a different paradigm on top of C++), or gcc, or evolution. Frankly, there are better languages for this.
Oh, and I'm not planning to use Ruby for the kernel or for Gcc, mind you. I'm not a specialist of Ruby, but the language looks quite appropriate for user interfaces or dynamic websites. If I had to choose a more modern language for the kernel, I would probably use either D, or OCaml, or a language with explicit continuations or explicit concurrency (Mozart-style). Some use C# and I'm willing to bet that their development will be significantly faster and that the resulting system is not going to be significantly slower.
I tend to assume that, after a few years of life and real-life profiling and optimization, a properly-designed system implemented in a high-level language would end up rather faster than a C kernel: while many low-level optimizations would have to be removed, a number of high-level optimizations can be introduced once you're sure of your code. Removing memory protection, for instance, is quite possible, if the system only executes programs without pointer arithmetics. In turn, this can be checked when the (possibly typed) binary is loaded, rather than at every single memory access. Let's not quabble about the fact that this check is hard-wired in the CPU. It's still a waste of time and energy somewhere and it's not the only one.

I was having trouble deciding whether he was kidding or what..then he put the muslim-communists punchline in and gave it away.

Dude:You should not be here. This is slashdot. Please go here.

or to stall the market until computer-based VOD can take over.

try doing a google search for "downloadable HD content" and see what you get. I'll give you a hint:

Microsoft's Todd Proebsting claims that compiler optimization only adds 4% performance per year, based on some back of the envelopes on x86 hardware.

This radical of a change in architecture should at least provide an accelerated growth from introduction through the next several years, which I'm sure will provide added incentive for those involved in compiler optimization -- finally, some real enhancements.

XP isn't faster then 2000. It apears faster in some situations but thats really a smoke screen(one you can benefit form though).

Both XP and 2000 are comparible in speed with 2000 being slightly faster at most things. Also you won't have to fix 2000 as much as XP when XP's smoke screens go bad.

(literaly i mean smokescreens. XP uses tricks to let you use the computer before it is completly booted and sometime it bites you.)

I appreciate your zest, but you are really off the game here.

I can assure you that the techs in my company's lab is very good at performance testing, and from a personal perspective, I can assure you XP is faster, although I should add a qualifier here. XP with SP2 is faster, RTM XP is only slightly faster than Win2k is some areas.

I know about the 'smoke' screens you are refering to, about the binding of the network after the Video intializes etc.

There are some true changes that are more of perception than actual performance, but rest assured this is NOT what I was talking about. (Also with SP2 & XP the binding orders changed again for security reasons, and a couple of the faster boot 'smoke' screens were removed.)

SP2 is more of a fork off the NT code branch after Windows 2003 Server, that is why it has better security as Windows 2003 was delayed for a major security shift at Microsoft. However, more importantly, code was further optimized in this process as well, and new compiler technology by Microsoft etc.

That is why when Windows 2003 Server shipped, a lot of geeks that had MSDN free access to it were running it on their desktops, it was faster than XP, even with all the extra Server 'services' running.

Then SP2 added in many of the changes of Windows 2003 to XP, and the performance honors went back to XP.

What makes XP faster has a lot to do with the timeline and SP2, but there are also a lot of kernel changes, allowing better caching, larger allocations of space in registry, better performance of loading in and out the registry, etc.

I don't have the links in front of me, but goto http://www.microsoft.com/ and do a search on: Mark Kernel WinXP

You will find the Microsoft documentation of the kernel changes between XP and Win2k, but you will also find an article written by Mark from SysInternals as well, that goes over the kernel differences from a non-Microsoft perspective.

These are part of the 'boost' XP has in running applications faster than Win2k. Also factor in the extra precautions for stability, the further optimizations in the Win32 subsystem and you get a faster OS, with more features.

Like I said before, our lab has 200mhz Pentium Laptops with 80mb of RAM and standard 4Gb Hard Drives. Swapping the drives with various OSes it is quick to see that WindowsXP is the fastext OS produced by Microsoft - it is even faster than Win98 that originally shipped on these laptops. (And no, these are not the only test systems our lab used to benchmark XP, but they are good example being low end hardware.)

I was also from the crowd that was, "No way, the themes, shadows, system restore, etc - XP just has to be slower." But it truly isn't...

I was blown away that the RTM version of XP was running neck and neck with Win2k, and SP2 just pulled further ahead. (I was amazed that Windows 2003 server was running faster as well before SP2 was released.)

I urge you to go do a bit of self research on this, and look past the articles that only focus on the smoke and mirror items, sure they are there, and MS fully admits they are there, but past them, there is some serious facts and basis to XP being faster.

Take Care...

You must have went to high school a long time ago. More seriously though, I believe that actually most schools, nonprofits, etc., have PCs that at least meet the following specifications that I pretty much chose at random because they provide an idea as to what is necessary to run an older Windows OS. All the ones (older Windows OS's) mentioned will perform e-mail functions without much difficulty. (Granted, I recognize that Microsoft, in that era, seriously understated the minimum requirements for their OS's -- but nevertheless, they are in the ballpark.) System Requirements* The following are the minimum system requirements to install and run Microsoft NetMeeting. * 90 megahertz (MHz) Pentium processor * 16 megabytes (MB) of RAM for Microsoft Windows 95, Windows 98, Windows Me * 24 megabytes (MB) of RAM for Microsoft Windows NT version 4.0 (Microsoft Windows NT 4.0 Service Pack 3 or later is required to enable sharing programs on Windows NT.) * Microsoft Internet Explorer version 4.01 or later * 28,800 bps or faster modem, integrated services digital network (ISDN), or local area network (LAN) connection (a fast Internet connection works best). * 4 MB of free hard disk space (an additional 10 MB is needed during installation only to accommodate the initial setup files). * Sound card with microphone and speakers (required for audio support). To use the data, audio, and video features of NetMeeting, your computer must meet the following hardware requirements: * For Windows 95, Windows 98, or Windows Me, a Pentium 90 processor with 16 MB of RAM (a Pentium 133 processor or better with at least 16 MB of RAM is recommended). * For Windows NT, a Pentium 90 processor with 24 MB of RAM (a Pentium 133 processor or better with at least 32 MB of RAM is recommended). * 4 MB of free hard disk space (an additional 10 MB is needed during installation only to accommodate the initial setup files). * 56,000 bps or faster modem, ISDN, or LAN connection. * Sound card with microphone and speakers (sound card required for both audio and video support). * Video capture card or camera that provides a Video for Windows capture driver (required for video support). *Source: http://www.microsoft.com/windows/NetMeeting/System Requirements/default.ASP The point of all this is that I do not believe that PCs with insufficient hardware specifications to run Windows 98 are much in use in this country. Consequently, underpowered PCs in need of an alternative (read in Linux) operating system adequate just for e-mail purposes are neither prevalent nor needed. Anecdotally, my mother is neither affluent nor technically savvy. She does uses an *old* PC that her children bought her that came preinstalled with Windows 98. It has since been upgraded with some additional RAM. She uses a cut-rate ISP -- Juno (that she is quite satisfied with) -- and as far as I can tell only uses the machine for e-mail. For her purposes, it is vastly overpowered. Over a period of five years there is probably $600 into it. Nobody in our family can think of a reason to upgrade it any further. I would like to have your description of a PC that cannot check e-mail because it is too underpowered to run any version of Windows/e-mail implementation, but an unsophisticated person with such a PC would benefit from a thin Linux installation on it. Perhaps you haven't noticed, but absolutely no one, except for perhaps a child geek, chooses Linux as their first OS. You are simply wrong when you say that you can just install Debian Linux and "tweak it a bit" and train a novice user. That is just complete Linux-will-save-the-world fantasy. "I think it's far more likely that more and more users will refuse to upgrade as they find Linux working on their old hardware." This is so extremely unlikely I will not argue the point. If anyone cares enough to reply with any supporting evidence whatsoever, I will be happy to d

It's hard to follow without the context of the original Statement of Objections to compare against. Does anyone know of a groklaw like website that is capturing all the publicly available legal documents on this case and organizing them in a coherent manner?

The message I get when reading it is that Microsoft says we're trying to comply but the commission is not cooperating and is biased against us and preventing us from bringing this to the Court of First Instance ("CFI") in Luxembourg.

Here are some of the more interesting snippets taken from Microsoft's document:

• 1. Hundreds of Microsoft employees and contractors have worked for more than 30,000 hours to create over 12,000 pages of detailed technical documents that are available for license today. In addition, Microsoft has offered to provide licensees with 500 hours of technical support and has made its source code related to all the relevant technologies available under a reference license.
• 3. The Commission continually changed its interpretation of what technical documentation was required by the vague language in the Decision, and refused to put its new interpretations in writing despite repeated requests from Microsoft [...]
• 4. The Commission has denied Microsoft's fundamental right of defence by prohibiting fair and full access to the file underlying the Statement of Objections, including correspondence between the Commission and the outside experts upon whose evidence the Commission relies.
• 9. The Commission refused to provide Microsoft a meaningful written statement, apparently because it wished to demand a broad scope of documentation, while, at the same time, preventing Microsoft from placing that fact squarely before the CFI in the appeal against the 2004 Decision (as substantiated by statements discussed in later sections of this Response). Microsoft, although dismayed by this gamesmanship, itself stated in writing that it would supply what it understood the Commission was requesting.
• 15. The Commission did not comment upon the 11 December 2004 draft documentation for more than six months, and then ignored the 8 August 2005 revision for several months more. Specifically, it never challenged Microsoft's description of the scope of the documentation that was being developed and supplied to the Commission.
• 18. With regard to the scope of Microsoft's Technical Documentation, the Statement of Objections claims that Microsoft has provided only "on-the-wire" protocol information, that is, information relating to how the protocols communicate information between computers in a Windows network, such as how data must be formatted by the sender to be read by the recipient, and how the meaning of the information transmitted can be understood. The Statement of Objections asserts that Microsoft has refused to supply a broader range of information which would help explain why the computers in a network communicate particular information and how the communicated information is used and with what results.
• 20. The usability problems asserted by the Statement of Objections relate to its ease of use. According to the Statement of Objections, descriptions of the proper sequencing of messages communicated between servers are not provided in a way "consistent with the kind of description commonly used in the industry" and in some instances are not provided at all.
• 24. The Commission has also contested the significance of Microsoft's voluntary offer to allow licensees to use the actual source code for Windows, even though the Commission itself demanded that the Trustee must be given the same code in order to determine Microsoft's compliance.
• 28. The Commission cannot have it both ways. If it claims

Win2003 requires,
- 133mhz processor
- 128mb of ram
- 1.25gb+ of hard drive space

From memory, that's a computer in the early 90's with some extra memory and a bigger hard drive, neither of which are anywhere near expensive.

It's no surprise that other server operating systems run on old hardware as well.

It's no surprise that Linux will run on older hardware,

I'm involved with a project that is looking to develop an online community for technology oriented business customers.
Sell your idea to ebay, they might like you. (and the highest bidder wins!)


  If you could develop an online community to encourage collaboration and information sharing, what features would you want included?
That's easy, BitTorrent.


  How would you go about including features that are widely available in other places (weblogging, message boards, wiki) and generating buy-in from customers.
1) Visit homepages of said OSS
2) Get the sources
3) Right-Click Ctrl-V
4) Get headache integrating code from multiple projects^W4) Discover 'magical' missing libraries^W4) Consider rewriting everything with existing code as reference^W4) Give up^W4) ????
5) Profit!

Are you saying that having only version 2.0 of the framework installed, I can run apps developed for .Net 1.0, 1.1 and 2.0? Or does it require to have multiple versions of the framework?

You can do either; its your choice.

And what if I have a 1.0 or 1.1 application and want to take advantage of one new feature in 2.0 in one (and only one) of my classes? Why should I have to convert the entire application?

I doubt there's one new feature that you'd want to use in just one class. That's just a silly thing to ask. BTW, the 'conversion' is usually just converting the project files and may some config files (if you tied your app to a specific version, for example, it will undo that.. why convert if you are only going to support the older runtime?). There's also a list of breaking changes, so you'll be able to easily 'fix' any code.

You might call it bashing, but it is not. Backward compatibility is a very important requirement, almost as much important as the feature set. It makes evolving a platform harder, but it can be done. Microsoft took the easy path, at expense of its customers. That is easy to do when you have vendor lock-in.

No, its bashing. Much of the changes are new additions to the framework and you have a list of breaking changes. If you want to take advantage of the new features you'll need to port. If you don't then leave it. Just like people that want to take advantage of the newest directx will need to target that library, or the newest java. Much of the framework is untouched from .Net 1.1.

Not wanting to sound as a fan boy here, but Java is a great counter-example - I don't know of any deprecated API that had actually been removed or stopped working.

That doesn't happen with moving to .Net 2 either. Look up the ObsoleleAttribute. There's nothign Obsoleted that will cause a compile error. You'll have plenty of time before those changes come.

"The Origami device is approximately 7 ½ inches x 3 1/2 inches x 1 1/2 inches thick and uses:
  Extensive multimedia features in Windows XP Embedded to support its customized video record and snapshot application and to provide video-conferencing capabilities
  Microsoft Internet Explorer 6 to provide the latest Web browsing capabilities including playback support for Flash and Shockwave files
  Outlook® Express for e-mail "

"At first glance, the device looks like a small camera and camcorder until you open it to reveal a brilliant 4-inch, 640 x 480 display that looks more like a terminal. "That's why we call the device Origami; it changes shape depending on what function you're using," Matson notes. "You twist it one way to use it as a personal access device, another way to write an e-mail, and other way to use it as a camera." A patent-pending hinge on the unit allows users to bend the device into different shapes."

http://members.microsoft.com/CustomerEvidence/Comm on/FileOpen.aspx?FileName=8998_National_Semiconduc tor_Conceptual_P.doc

This is a detailled answer to the criticism by the EU. It looks as if the EU complains that the new API documentation doesn't explain what a handle is or what exactly a void pointer points to.

My impression is that the person who reviewed the documentation on the EU side has no clue about programming.

Actually, every single 'hidden' API that is used outside of windows core code is documented here. In addition to that, Microsoft has made the source code public, given full details on implementation (you have to submit a request in writing and sign a form about not copying it, though).

Now, they won't make your program magically run faster, they are just some APIs that people used for convenience when they shouldn't have.

The problem with the current state of the case is that the EU demands more, without specifying what more they would like to see.

My guess? They are simply after more money.

REDMOND, Wash., Feb. 23, 2006 - Microsoft is making available its formal response to the Statement of Objections issued in December 2005 by the European Commission.

The response, which was filed Feb. 15 in Brussels, details the evidence that Microsoft is in full compliance with the technical documentation requirements imposed by the Commission in 2004. It also details numerous ways in which the Commission had ignored key information and denied Microsoft due process in defending itself.

Microsoft also is releasing two independent expert reports by software system engineering professors who examined the technical documentation created by Microsoft.

Cover letter (.PDF file, 2 pages, 160kb)
Microsoft's Response to the Commission's Statement of Objections (.PDF file, 78 pages, 800 kb)
Annex 1: Compliance Effort Report (.PDF file, 5 pages, 116 kb)
Annex 2: MCCP Report (.PDF file, 13 pages, 177kb)
Annex 3: Finkelstein Report (.PDF file, 654kb)
Annex 4: Broy Report (.PDF file, 42 pages, 570kb)
Annex 5: Hirst Report (.PDF file, 13 pages, 1.7mb)
Annex 6: Thatcher Report (.PDF file, 6 pages, 116kb)
Annex 7: Criticisms Report (.PDF file, 17 pages, 200kb)
Exhibits to SO Response (.PDF file, 22 pages, 696kb)

I'm puzzled by this, but I don't know enough about servers and their protocols to be able to form an opinion about the reports. Could someone please read them and tell us if they exonerate Microsoft?

Purely from a formal point of view ... if the EU demands openness in the sense of interoperability, and Microsoft created an architecture that cannot be opened without copying the whole server. In that case things may look as bleak for Microsoft as it would if it turned out that it simply refused to comply.